Health Search and Machine-Learned Epidemiology

Date: 
December 5, 2018
Time: 
3:00 pm - 4:00 pm
Place: 
MH-2700

Evgenly Gabrilovich, Phd, Senior Staff Research Scientist at Google

Title: Health Search and Machine-Learned Epidemiology

Summary: Approximately 1 percent of all Google searches are symptom-related, as users are conducting online research on pertinent medical conditions. In the first part of this talk, we will discuss the symptom search experience on Google, where we use machine learning methods to identify medical conditions relevant to the constellation of symptoms in the query. Then, we will show how to harness the scale and the richness of aggregated web search and location data to build timely models of population health. We will present FINDER, a machine-learned model for real-time detection of foodborne illness. Our method identifies potentially unsafe restaurants by estimating the fraction of people who visited a particular restaurant and later searched for terms indicative of food poisoning. FINDER enables us to ascertain previously intractable epidemiological information, for example, in 38% of cases the restaurant potentially causing food poisoning was not the last one visited. We will conclude the talk with a discussion of how aggregate mobility data can help support relief efforts after natural disasters

Event Type: 
Biostatistics and Bioinformatics Seminar